The Command History page is the audit trail for every command queued in the system: scripts, diagnose runs, message popups, deploys, checks. It's filterable by device, type, status, date range, and free-text search across stdout/stderr.
What's captured
For every command:
- Status:
pending,running,completed,failed,timeout,cancelled - Stdout + stderr (PII-scrubbed at the agent before transmit)
- Exit code
- Started / completed timestamps
- Duration
- Streamed phase logs (for long-running types like
deploy_agent) - Audit log entry: who ran it, when, with what payload
Live log pane
Click any running command to expand a live-tailing log pane. Phase events (reached_target, download_installer, api_key_injection, etc.) stream in as the agent emits them. Polled every 1.5 s.
Timeout handling
Each command type has a per-type timeout. After the timeout, a background job marks the command timeout and stops the agent from honouring a stale result.
| Type | Timeout |
|---|---|
send_message |
30 s |
run_script |
5 min |
deploy_agent |
15 min |
diagnose |
20 min |
device_check |
60 s default (configurable per-check) |
Redaction badge
When a stdout/stderr block contains scrubber tokens (<USER>, <HOSTNAME>, <EMAIL>, etc.), a small "🛡 N items redacted locally" badge appears. The PII never reached our servers, see PII scrubber.